Ethan Mollick: GPT-6 Pro still has no equivalent — one-shot hard tasks with great communication
emollick · x · 2026-10-05
Wharton professor Ethan Mollick, despite complaining about some OpenAI choices, notes there is still no equivalent to GPT-6 Pro: like previous Pro models, it both completes genuinely hard tasks in one shot and communicates the results surprisingly well. A rare positive take from a high-profile AI commentator.
More from Models
- Mobilerun scores perfect 116/116 on Google's AndroidWorld, beating Artemis — SimplyAnnisa · 2026-10-05
- Weekend iOS app + pSEO project burned through full Opus quota plus ~$450 across 7 threads — iannuttall · 2026-10-05
- Same system scores 51.4 vs 75.8 depending on judge — are memory benchmarks trustworthy? — Efficient_Joke3384 · 2026-10-05
- AWS to unplug Claude Sonnet 3.x on July 30, sparking emotional user farewell drama — repligate · 2026-10-05
- GPTs' "demanding" persona traced to stable value vectors since GPT-5, argue devs — repligate · 2026-10-05
- Redditor suspects Claude usage drains faster during US peak hours despite Anthropic's May change — Mousse_External · 2026-10-05